IS MORE EVER TOO MUCH - THE NUMBER OF INDICATORS PER FACTOR IN CONFIRMATORY FACTOR-ANALYSIS

We evaluated whether ''more is ever too much'' for the number of indic ators (p) per factor (p/f) in confirmatory factor analysis by varying sample size (N = 50-1000) and p/f(2-12 items per factor) in 35,000 Mon te Carlo solutions. For all Ns, solution behavior steadily improved (m ore proper solutions, more accurate parameter estimates; greater relia bility) with increasing p/f: There was a compensatory relation between N and p/f: large p/f compensated for small Nand large N compensated f or small p/f but large-N and large-p/f was best. A bias in the behavio r of the chi(2) was also demonstrated where apparent goodness of fit d eclined with increasing p/f ratios even though approximating models we re ''true''. Fit was similar for proper and improper solutions, as wer e parameter estimates from improper solutions not involving offending estimates. We also used the 12-p/f data to construct 2, 3, 4, or 6 par cels of items (e.g., two parcels of 6 items per factor, three parcels of 4 items per factor, etc.), but the 12-indicator (nonparceled) solut ions were somewhat better behaved. At least for conditions in our simu lation study, traditional ''rules'' implying fewer indicators should b e used for smaller N may be inappropriate and researchers should consi der using more indicators per factor than is evident in current practi ce.